diffusers 0.24.0__py3-none-any.whl → 0.25.1__py3-none-any.whl

diffusers/schedulers/scheduling_euler_discrete.py

@@ -92,6 +92,43 @@ def betas_for_alpha_bar(
     return torch.tensor(betas, dtype=torch.float32)
 
 
+# Copied from diffusers.schedulers.scheduling_ddim.rescale_zero_terminal_snr
+def rescale_zero_terminal_snr(betas):
+    """
+    Rescales betas to have zero terminal SNR Based on https://arxiv.org/pdf/2305.08891.pdf (Algorithm 1)
+
+
+    Args:
+        betas (`torch.FloatTensor`):
+            the betas that the scheduler is being initialized with.
+
+    Returns:
+        `torch.FloatTensor`: rescaled betas with zero terminal SNR
+    """
+    # Convert betas to alphas_bar_sqrt
+    alphas = 1.0 - betas
+    alphas_cumprod = torch.cumprod(alphas, dim=0)
+    alphas_bar_sqrt = alphas_cumprod.sqrt()
+
+    # Store old values.
+    alphas_bar_sqrt_0 = alphas_bar_sqrt[0].clone()
+    alphas_bar_sqrt_T = alphas_bar_sqrt[-1].clone()
+
+    # Shift so the last timestep is zero.
+    alphas_bar_sqrt -= alphas_bar_sqrt_T
+
+    # Scale so the first timestep is back to the old value.
+    alphas_bar_sqrt *= alphas_bar_sqrt_0 / (alphas_bar_sqrt_0 - alphas_bar_sqrt_T)
+
+    # Convert alphas_bar_sqrt to betas
+    alphas_bar = alphas_bar_sqrt**2  # Revert sqrt
+    alphas = alphas_bar[1:] / alphas_bar[:-1]  # Revert cumprod
+    alphas = torch.cat([alphas_bar[0:1], alphas])
+    betas = 1 - alphas
+
+    return betas
+
+
 class EulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
     """
     Euler scheduler.
@@ -128,6 +165,10 @@ class EulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
             An offset added to the inference steps. You can use a combination of `offset=1` and
             `set_alpha_to_one=False` to make the last step use step 0 for the previous alpha product like in Stable
             Diffusion.
+        rescale_betas_zero_snr (`bool`, defaults to `False`):
+            Whether to rescale the betas to have zero terminal SNR. This enables the model to generate very bright and
+            dark samples instead of limiting it to samples with medium brightness. Loosely related to
+            [`--offset_noise`](https://github.com/huggingface/diffusers/blob/74fd735eb073eb1d774b1ab4154a0876eb82f055/examples/dreambooth/train_dreambooth.py#L506).
     """
 
     _compatibles = [e.name for e in KarrasDiffusionSchedulers]
@@ -149,6 +190,7 @@ class EulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
         timestep_spacing: str = "linspace",
         timestep_type: str = "discrete",  # can be "discrete" or "continuous"
         steps_offset: int = 0,
+        rescale_betas_zero_snr: bool = False,
     ):
         if trained_betas is not None:
             self.betas = torch.tensor(trained_betas, dtype=torch.float32)
@@ -163,9 +205,17 @@ class EulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
         else:
             raise NotImplementedError(f"{beta_schedule} does is not implemented for {self.__class__}")
 
+        if rescale_betas_zero_snr:
+            self.betas = rescale_zero_terminal_snr(self.betas)
+
         self.alphas = 1.0 - self.betas
         self.alphas_cumprod = torch.cumprod(self.alphas, dim=0)
 
+        if rescale_betas_zero_snr:
+            # Close to 0 without being 0 so first sigma is not inf
+            # FP16 smallest positive subnormal works well here
+            self.alphas_cumprod[-1] = 2**-24
+
         sigmas = np.array(((1 - self.alphas_cumprod) / self.alphas_cumprod) ** 0.5)
         timesteps = np.linspace(0, num_train_timesteps - 1, num_train_timesteps, dtype=float)[::-1].copy()
 
@@ -187,14 +237,16 @@ class EulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
         self.use_karras_sigmas = use_karras_sigmas
 
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     @property
     def init_noise_sigma(self):
         # standard deviation of the initial noise distribution
+        max_sigma = max(self.sigmas) if isinstance(self.sigmas, list) else self.sigmas.max()
         if self.config.timestep_spacing in ["linspace", "trailing"]:
-            return self.sigmas.max()
+            return max_sigma
 
-        return (self.sigmas.max() ** 2 + 1) ** 0.5
+        return (max_sigma**2 + 1) ** 0.5
 
     @property
     def step_index(self):
@@ -269,7 +321,7 @@ class EulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
         if self.config.interpolation_type == "linear":
             sigmas = np.interp(timesteps, np.arange(0, len(sigmas)), sigmas)
         elif self.config.interpolation_type == "log_linear":
-            sigmas = torch.linspace(np.log(sigmas[-1]), np.log(sigmas[0]), num_inference_steps + 1).exp()
+            sigmas = torch.linspace(np.log(sigmas[-1]), np.log(sigmas[0]), num_inference_steps + 1).exp().numpy()
         else:
             raise ValueError(
                 f"{self.config.interpolation_type} is not implemented. Please specify interpolation_type to either"
@@ -290,6 +342,7 @@ class EulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
 
         self.sigmas = torch.cat([sigmas, torch.zeros(1, device=sigmas.device)])
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     def _sigma_to_t(self, sigma, log_sigmas):
         # get log sigma
@@ -419,6 +472,9 @@ class EulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
         if self.step_index is None:
             self._init_step_index(timestep)
 
+        # Upcast to avoid precision issues when computing prev_sample
+        sample = sample.to(torch.float32)
+
         sigma = self.sigmas[self.step_index]
 
         gamma = min(s_churn / (len(self.sigmas) - 1), 2**0.5 - 1) if s_tmin <= sigma <= s_tmax else 0.0
@@ -455,6 +511,9 @@ class EulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
 
         prev_sample = sample + derivative * dt
 
+        # Cast sample back to model compatible dtype
+        prev_sample = prev_sample.to(model_output.dtype)
+
         # upon completion increase step index by one
         self._step_index += 1
 

diffusers/schedulers/scheduling_heun_discrete.py

@@ -148,6 +148,7 @@ class HeunDiscreteScheduler(SchedulerMixin, ConfigMixin):
         self.use_karras_sigmas = use_karras_sigmas
 
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     def index_for_timestep(self, timestep, schedule_timesteps=None):
         if schedule_timesteps is None:
@@ -269,6 +270,7 @@ class HeunDiscreteScheduler(SchedulerMixin, ConfigMixin):
         self.dt = None
 
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
         # (YiYi Notes: keep this for now since we are keeping add_noise function which use index_for_timestep)
         # for exp beta schedules, such as the one for `pipeline_shap_e.py`

diffusers/schedulers/scheduling_k_dpm_2_ancestral_discrete.py

@@ -140,6 +140,7 @@ class KDPM2AncestralDiscreteScheduler(SchedulerMixin, ConfigMixin):
         #  set all values
         self.set_timesteps(num_train_timesteps, None, num_train_timesteps)
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     # Copied from diffusers.schedulers.scheduling_heun_discrete.HeunDiscreteScheduler.index_for_timestep
     def index_for_timestep(self, timestep, schedule_timesteps=None):
@@ -295,6 +296,7 @@ class KDPM2AncestralDiscreteScheduler(SchedulerMixin, ConfigMixin):
         self._index_counter = defaultdict(int)
 
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     # Copied from diffusers.schedulers.scheduling_euler_discrete.EulerDiscreteScheduler._sigma_to_t
     def _sigma_to_t(self, sigma, log_sigmas):

diffusers/schedulers/scheduling_k_dpm_2_discrete.py

@@ -140,6 +140,7 @@ class KDPM2DiscreteScheduler(SchedulerMixin, ConfigMixin):
         self.set_timesteps(num_train_timesteps, None, num_train_timesteps)
 
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     # Copied from diffusers.schedulers.scheduling_heun_discrete.HeunDiscreteScheduler.index_for_timestep
     def index_for_timestep(self, timestep, schedule_timesteps=None):
@@ -284,6 +285,7 @@ class KDPM2DiscreteScheduler(SchedulerMixin, ConfigMixin):
         self._index_counter = defaultdict(int)
 
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     @property
     def state_in_first_order(self):

diffusers/schedulers/scheduling_lms_discrete.py

@@ -168,6 +168,7 @@ class LMSDiscreteScheduler(SchedulerMixin, ConfigMixin):
         self.is_scale_input_called = False
 
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     @property
     def init_noise_sigma(self):
@@ -279,6 +280,7 @@ class LMSDiscreteScheduler(SchedulerMixin, ConfigMixin):
         self.sigmas = torch.from_numpy(sigmas).to(device=device)
         self.timesteps = torch.from_numpy(timesteps).to(device=device)
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
         self.derivatives = []
 

diffusers/schedulers/scheduling_unipc_multistep.py

@@ -175,6 +175,7 @@ class UniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
         self.alpha_t = torch.sqrt(self.alphas_cumprod)
         self.sigma_t = torch.sqrt(1 - self.alphas_cumprod)
         self.lambda_t = torch.log(self.alpha_t) - torch.log(self.sigma_t)
+        self.sigmas = ((1 - self.alphas_cumprod) / self.alphas_cumprod) ** 0.5
 
         # standard deviation of the initial noise distribution
         self.init_noise_sigma = 1.0
@@ -197,6 +198,7 @@ class UniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
         self.solver_p = solver_p
         self.last_sample = None
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     @property
     def step_index(self):
@@ -267,6 +269,7 @@ class UniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
 
         # add an index counter for schedulers that allow duplicated timesteps
         self._step_index = None
+        self.sigmas.to("cpu")  # to avoid too much CPU/GPU communication
 
     # Copied from diffusers.schedulers.scheduling_ddpm.DDPMScheduler._threshold_sample
     def _threshold_sample(self, sample: torch.FloatTensor) -> torch.FloatTensor:
@@ -827,7 +830,16 @@ class UniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
             schedule_timesteps = self.timesteps.to(original_samples.device)
             timesteps = timesteps.to(original_samples.device)
 
-        step_indices = [(schedule_timesteps == t).nonzero().item() for t in timesteps]
+        step_indices = []
+        for timestep in timesteps:
+            index_candidates = (schedule_timesteps == timestep).nonzero()
+            if len(index_candidates) == 0:
+                step_index = len(schedule_timesteps) - 1
+            elif len(index_candidates) > 1:
+                step_index = index_candidates[1].item()
+            else:
+                step_index = index_candidates[0].item()
+            step_indices.append(step_index)
 
         sigma = sigmas[step_indices].flatten()
         while len(sigma.shape) < len(original_samples.shape):

diffusers/schedulers/scheduling_utils.py

@@ -18,6 +18,7 @@ from enum import Enum
 from typing import Optional, Union
 
 import torch
+from huggingface_hub.utils import validate_hf_hub_args
 
 from ..utils import BaseOutput, PushToHubMixin
 
@@ -81,6 +82,7 @@ class SchedulerMixin(PushToHubMixin):
     has_compatibles = True
 
     @classmethod
+    @validate_hf_hub_args
     def from_pretrained(
         cls,
         pretrained_model_name_or_path: Optional[Union[str, os.PathLike]] = None,
@@ -120,7 +122,7 @@ class SchedulerMixin(PushToHubMixin):
             local_files_only(`bool`, *optional*, defaults to `False`):
                 Whether to only load local model weights and configuration files or not. If set to `True`, the model
                 won't be downloaded from the Hub.
-            use_auth_token (`str` or *bool*, *optional*):
+            token (`str` or *bool*, *optional*):
                 The token to use as HTTP bearer authorization for remote files. If `True`, the token generated from
                 `diffusers-cli login` (stored in `~/.huggingface`) is used.
             revision (`str`, *optional*, defaults to `"main"`):

diffusers/schedulers/scheduling_utils_flax.py

@@ -20,6 +20,7 @@ from typing import Optional, Tuple, Union
 
 import flax
 import jax.numpy as jnp
+from huggingface_hub.utils import validate_hf_hub_args
 
 from ..utils import BaseOutput, PushToHubMixin
 
@@ -70,6 +71,7 @@ class FlaxSchedulerMixin(PushToHubMixin):
     has_compatibles = True
 
     @classmethod
+    @validate_hf_hub_args
     def from_pretrained(
         cls,
         pretrained_model_name_or_path: Optional[Union[str, os.PathLike]] = None,
@@ -110,7 +112,7 @@ class FlaxSchedulerMixin(PushToHubMixin):
                 Whether or not to also return a dictionary containing missing keys, unexpected keys and error messages.
             local_files_only(`bool`, *optional*, defaults to `False`):
                 Whether or not to only look at local files (i.e., do not try to download the model).
-            use_auth_token (`str` or *bool*, *optional*):
+            token (`str` or *bool*, *optional*):
                 The token to use as HTTP bearer authorization for remote files. If `True`, will use the token generated
                 when running `transformers-cli login` (stored in `~/.huggingface`).
             revision (`str`, *optional*, defaults to `"main"`):

diffusers/training_utils.py

@@ -67,7 +67,7 @@ def unet_lora_state_dict(unet: UNet2DConditionModel) -> Dict[str, torch.Tensor]:
                 current_lora_layer_sd = lora_layer.state_dict()
                 for lora_layer_matrix_name, lora_param in current_lora_layer_sd.items():
                     # The matrix name can either be "down" or "up".
-                    lora_state_dict[f"unet.{name}.lora.{lora_layer_matrix_name}"] = lora_param
+                    lora_state_dict[f"{name}.lora.{lora_layer_matrix_name}"] = lora_param
 
     return lora_state_dict
 

diffusers/utils/__init__.py

@@ -21,7 +21,6 @@ from .. import __version__
 from .constants import (
     CONFIG_NAME,
     DEPRECATED_REVISION_ARGS,
-    DIFFUSERS_CACHE,
     DIFFUSERS_DYNAMIC_MODULE_NAME,
     FLAX_WEIGHTS_NAME,
     HF_MODULES_CACHE,
@@ -38,7 +37,6 @@ from .doc_utils import replace_example_docstring
 from .dynamic_modules_utils import get_class_from_dynamic_module
 from .export_utils import export_to_gif, export_to_obj, export_to_ply, export_to_video
 from .hub_utils import (
-    HF_HUB_OFFLINE,
     PushToHubMixin,
     _add_variant,
     _get_model_file,

diffusers/utils/constants.py

@@ -14,15 +14,13 @@
 import importlib
 import os
 
-from huggingface_hub.constants import HUGGINGFACE_HUB_CACHE, hf_cache_home
+from huggingface_hub.constants import HF_HOME
 from packaging import version
 
 from ..dependency_versions_check import dep_version_check
 from .import_utils import ENV_VARS_TRUE_VALUES, is_peft_available, is_transformers_available
 
 
-default_cache_path = HUGGINGFACE_HUB_CACHE
-
 MIN_PEFT_VERSION = "0.6.0"
 MIN_TRANSFORMERS_VERSION = "4.34.0"
 _CHECK_PEFT = os.environ.get("_CHECK_PEFT", "1") in ENV_VARS_TRUE_VALUES
@@ -35,9 +33,8 @@ ONNX_WEIGHTS_NAME = "model.onnx"
 SAFETENSORS_WEIGHTS_NAME = "diffusion_pytorch_model.safetensors"
 ONNX_EXTERNAL_WEIGHTS_NAME = "weights.pb"
 HUGGINGFACE_CO_RESOLVE_ENDPOINT = os.environ.get("HF_ENDPOINT", "https://huggingface.co")
-DIFFUSERS_CACHE = default_cache_path
 DIFFUSERS_DYNAMIC_MODULE_NAME = "diffusers_modules"
-HF_MODULES_CACHE = os.getenv("HF_MODULES_CACHE", os.path.join(hf_cache_home, "modules"))
+HF_MODULES_CACHE = os.getenv("HF_MODULES_CACHE", os.path.join(HF_HOME, "modules"))
 DEPRECATED_REVISION_ARGS = ["fp16", "non-ema"]
 
 # Below should be `True` if the current version of `peft` and `transformers` are compatible with

diffusers/utils/dummy_pt_objects.py

@@ -302,6 +302,21 @@ class UNetSpatioTemporalConditionModel(metaclass=DummyObject):
         requires_backends(cls, ["torch"])
 
 
+class UVit2DModel(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+    @classmethod
+    def from_config(cls, *args, **kwargs):
+        requires_backends(cls, ["torch"])
+
+    @classmethod
+    def from_pretrained(cls, *args, **kwargs):
+        requires_backends(cls, ["torch"])
+
+
 class VQModel(metaclass=DummyObject):
     _backends = ["torch"]
 
@@ -645,6 +660,21 @@ class ScoreSdeVePipeline(metaclass=DummyObject):
         requires_backends(cls, ["torch"])
 
 
+class AmusedScheduler(metaclass=DummyObject):
+    _backends = ["torch"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch"])
+
+    @classmethod
+    def from_config(cls, *args, **kwargs):
+        requires_backends(cls, ["torch"])
+
+    @classmethod
+    def from_pretrained(cls, *args, **kwargs):
+        requires_backends(cls, ["torch"])
+
+
 class CMStochasticIterativeScheduler(metaclass=DummyObject):
     _backends = ["torch"]
 

diffusers/utils/dummy_torch_and_transformers_objects.py

@@ -32,6 +32,51 @@ class AltDiffusionPipeline(metaclass=DummyObject):
         requires_backends(cls, ["torch", "transformers"])
 
 
+class AmusedImg2ImgPipeline(metaclass=DummyObject):
+    _backends = ["torch", "transformers"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch", "transformers"])
+
+    @classmethod
+    def from_config(cls, *args, **kwargs):
+        requires_backends(cls, ["torch", "transformers"])
+
+    @classmethod
+    def from_pretrained(cls, *args, **kwargs):
+        requires_backends(cls, ["torch", "transformers"])
+
+
+class AmusedInpaintPipeline(metaclass=DummyObject):
+    _backends = ["torch", "transformers"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch", "transformers"])
+
+    @classmethod
+    def from_config(cls, *args, **kwargs):
+        requires_backends(cls, ["torch", "transformers"])
+
+    @classmethod
+    def from_pretrained(cls, *args, **kwargs):
+        requires_backends(cls, ["torch", "transformers"])
+
+
+class AmusedPipeline(metaclass=DummyObject):
+    _backends = ["torch", "transformers"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["torch", "transformers"])
+
+    @classmethod
+    def from_config(cls, *args, **kwargs):
+        requires_backends(cls, ["torch", "transformers"])
+
+    @classmethod
+    def from_pretrained(cls, *args, **kwargs):
+        requires_backends(cls, ["torch", "transformers"])
+
+
 class AnimateDiffPipeline(metaclass=DummyObject):
     _backends = ["torch", "transformers"]
 

diffusers/utils/dynamic_modules_utils.py

@@ -25,7 +25,8 @@ from pathlib import Path
 from typing import Dict, Optional, Union
 from urllib import request
 
-from huggingface_hub import HfFolder, cached_download, hf_hub_download, model_info
+from huggingface_hub import cached_download, hf_hub_download, model_info
+from huggingface_hub.utils import validate_hf_hub_args
 from packaging import version
 
 from .. import __version__
@@ -194,6 +195,7 @@ def find_pipeline_class(loaded_module):
     return pipeline_class
 
 
+@validate_hf_hub_args
 def get_cached_module_file(
     pretrained_model_name_or_path: Union[str, os.PathLike],
     module_file: str,
@@ -201,7 +203,7 @@ def get_cached_module_file(
     force_download: bool = False,
     resume_download: bool = False,
     proxies: Optional[Dict[str, str]] = None,
-    use_auth_token: Optional[Union[bool, str]] = None,
+    token: Optional[Union[bool, str]] = None,
     revision: Optional[str] = None,
     local_files_only: bool = False,
 ):
@@ -232,7 +234,7 @@ def get_cached_module_file(
         proxies (`Dict[str, str]`, *optional*):
             A dictionary of proxy servers to use by protocol or endpoint, e.g., `{'http': 'foo.bar:3128',
             'http://hostname': 'foo.bar:4012'}.` The proxies are used on each request.
-        use_auth_token (`str` or *bool*, *optional*):
+        token (`str` or *bool*, *optional*):
             The token to use as HTTP bearer authorization for remote files. If `True`, will use the token generated
             when running `transformers-cli login` (stored in `~/.huggingface`).
         revision (`str`, *optional*, defaults to `"main"`):
@@ -244,7 +246,7 @@ def get_cached_module_file(
 
     <Tip>
 
-    You may pass a token in `use_auth_token` if you are not logged in (`huggingface-cli long`) and want to use private
+    You may pass a token in `token` if you are not logged in (`huggingface-cli login`) and want to use private
     or [gated models](https://huggingface.co/docs/hub/models-gated#gated-models).
 
     </Tip>
@@ -289,7 +291,7 @@ def get_cached_module_file(
                 proxies=proxies,
                 resume_download=resume_download,
                 local_files_only=local_files_only,
-                use_auth_token=False,
+                token=False,
             )
             submodule = "git"
             module_file = pretrained_model_name_or_path + ".py"
@@ -307,7 +309,7 @@ def get_cached_module_file(
                 proxies=proxies,
                 resume_download=resume_download,
                 local_files_only=local_files_only,
-                use_auth_token=use_auth_token,
+                token=token,
             )
             submodule = os.path.join("local", "--".join(pretrained_model_name_or_path.split("/")))
         except EnvironmentError:
@@ -332,13 +334,6 @@ def get_cached_module_file(
     else:
         # Get the commit hash
         # TODO: we will get this info in the etag soon, so retrieve it from there and not here.
-        if isinstance(use_auth_token, str):
-            token = use_auth_token
-        elif use_auth_token is True:
-            token = HfFolder.get_token()
-        else:
-            token = None
-
         commit_hash = model_info(pretrained_model_name_or_path, revision=revision, token=token).sha
 
         # The module file will end up being placed in a subfolder with the git hash of the repo. This way we get the
@@ -359,13 +354,14 @@ def get_cached_module_file(
                     force_download=force_download,
                     resume_download=resume_download,
                     proxies=proxies,
-                    use_auth_token=use_auth_token,
+                    token=token,
                     revision=revision,
                     local_files_only=local_files_only,
                 )
     return os.path.join(full_submodule, module_file)
 
 
+@validate_hf_hub_args
 def get_class_from_dynamic_module(
     pretrained_model_name_or_path: Union[str, os.PathLike],
     module_file: str,
@@ -374,7 +370,7 @@ def get_class_from_dynamic_module(
     force_download: bool = False,
     resume_download: bool = False,
     proxies: Optional[Dict[str, str]] = None,
-    use_auth_token: Optional[Union[bool, str]] = None,
+    token: Optional[Union[bool, str]] = None,
     revision: Optional[str] = None,
     local_files_only: bool = False,
     **kwargs,
@@ -414,7 +410,7 @@ def get_class_from_dynamic_module(
         proxies (`Dict[str, str]`, *optional*):
             A dictionary of proxy servers to use by protocol or endpoint, e.g., `{'http': 'foo.bar:3128',
             'http://hostname': 'foo.bar:4012'}.` The proxies are used on each request.
-        use_auth_token (`str` or `bool`, *optional*):
+        token (`str` or `bool`, *optional*):
             The token to use as HTTP bearer authorization for remote files. If `True`, will use the token generated
             when running `transformers-cli login` (stored in `~/.huggingface`).
         revision (`str`, *optional*, defaults to `"main"`):
@@ -426,7 +422,7 @@ def get_class_from_dynamic_module(
 
     <Tip>
 
-    You may pass a token in `use_auth_token` if you are not logged in (`huggingface-cli long`) and want to use private
+    You may pass a token in `token` if you are not logged in (`huggingface-cli login`) and want to use private
     or [gated models](https://huggingface.co/docs/hub/models-gated#gated-models).
 
     </Tip>
@@ -449,7 +445,7 @@ def get_class_from_dynamic_module(
         force_download=force_download,
         resume_download=resume_download,
         proxies=proxies,
-        use_auth_token=use_auth_token,
+        token=token,
         revision=revision,
         local_files_only=local_files_only,
     )